[ROOFLINE] Add CUDA support to roofline analysis #12205
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
|
Will take a look tomorrow |
Contributor
AndrewZhaoLuo
left a comment
AndrewZhaoLuo
left a comment
There was a problem hiding this comment.
Need to grok the tensorcore stuff a bit but seems good so far. On my 3070
I get 420 Gb/s bandwidth vs the 448 advertised. For the TFLops I actually get more than the 40.6 TFLops advertised (I get 41.2 TFlops which seems close enough)
added 6 commits
July 29, 2022 08:56
Add functions to estimate peak flops and bandwidth for CUDA. Add a new registration mechanism to the roofline analysis to support adding any target. This mechanism uses generic functions with overrides. New targets only need to add `estimate_peak_bandwidth` and `estimate_peak_flops` functions. Also fix cuda codegen and tensorcore_infer_fragment.cc to support filling matrix_a and matrix_b fragments.
AndrewZhaoLuo
approved these changes
Jul 30, 2022
xinetzone
pushed a commit
to daobook/tvm
that referenced
this pull request
Nov 25, 2022
* [ROOFLINE] Add CUDA support to roofline analysis Add functions to estimate peak flops and bandwidth for CUDA. Add a new registration mechanism to the roofline analysis to support adding any target. This mechanism uses generic functions with overrides. New targets only need to add `estimate_peak_bandwidth` and `estimate_peak_flops` functions. Also fix cuda codegen and tensorcore_infer_fragment.cc to support filling matrix_a and matrix_b fragments. * formatiing * move statement back inside loops * print out report for debugging * default to avx2 * review comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add functions to estimate peak flops and bandwidth for CUDA. Add a new registration mechanism to the roofline analysis to support adding any target. This mechanism uses generic functions with overrides. New targets only need to add
estimate_peak_bandwidthandestimate_peak_flopsfunctions.Also fix cuda codegen and tensorcore_infer_fragment.cc to support filling matrix_a and matrix_b fragments.
@AndrewZhaoLuo